CAPILLARY SHRINKAGE CRACKING AND ITS PREVENTION BY

CONTROLLED CONCRETE CURING

Markus Schmidt (1, 2) and Volker Slowik (1)

(1) Leipzig University of Applied Sciences, Leipzig, Germany

(2) University of the West of Scotland, Paisley, UK

Abstract
The build-up of a negative pressure in the liquid phase of a drying suspension may lead to
cracking. In concrete construction, this effect results in damage processes taking place already
in the very early age, i.e. within the first few hours after casting when the concrete has not yet
reached a significant strength. For avoiding this type of damage, a method of controlled
concrete curing has been proposed. It is based on in situ capillary pressure measurement and
closed-loop controlled rewetting of fresh concrete surfaces. The capillary pressure is kept
below a critical value and an uncontrolled application of too much water to the concrete
surface is prevented. The method has been tested under laboratory as well as under site
conditions.

1. INTRODUCTION
Under high evaporation rates, mainly caused by wind, low relative air humidity and high
temperature, concrete may crack even before the material has reached a significant strength.
Figure 1 shows a concrete slab which has been cast on a sunny and windy day. Since the
curing was not started at the right time, cracks were formed within the first four hours after
casting. These cracks had widths of about 1 mm and large depths. Some of them split the
whole structure.
The process which leads to this type of early age cracking is the so-called plastic or
capillary shrinkage. Water loss causes the build-up of a negative capillary pressure in the
water filled pore system of the plastic material. Especially in high performance concrete with
low water-binder-ratios, the capillary pressure development in the very early age is also
affected by self-desiccation. If the capillary shrinkage is hindered cracks may occur.
In many cases, early age cracks are not as large and visible as shown in Figure 1.
Sometimes these cracks are very small or they are temporarily covered during surface
finishing. Nevertheless, they might have an effect on the structural durability. Numerical
simulations have shown that early age damage may influence drying shrinkage cracking of the
hardened concrete [1] as well as the cracking under the action of external forces.

Page 1
 Figure 1: Cracks in a concrete slab caused by capillary shrinkage

2. CAPILLARY PRESSURE DEVELOPMENT

After the placing of the concrete, its surface is normally covered by a thin film of water
(Figure 2, A). Consolidation of the solid particles contributes into the accumulation of water
at the concrete surface. This effect is called bleeding. When water on the surface evaporates
and the surface is not completely covered by a plane water film anymore, menisci are formed
between the solid particles, see Figure 2, B.

Figure 2: Capillary pressure build-up

The formation of the menisci results in a negative capillary pressure in the water filled
pores. This physical process is described by the Gauss-Laplace-Equation:
 1 1 
p = − γ  +  (1)
R
 1 R2 

with p = capillary pressure, γ = surface tension of the fluid and R1, R2 = main radii of the
curvilinear fluid surface.
The capillary pressure is inversely proportional to the radii of the menisci. The smaller the
spaces between the particles, the smaller the radii of the menisci and the higher the absolute

Page 2
capillary pressure value can become. Since the pressure acts on the solid particles of the
drying suspension it results in a contraction of the material. Hence, the pores formed by the
solid particles are becoming smaller (Figure 2, C). During this process, water is drawn out of
the pores and transported to the surface where it evaporates. This happens until the particles
can not get closer anymore due to restrictions like contact and friction or due to chemical
processes like hydration. If not enough water is transported to the surface and the radii of the
menisci become as small as the pore diameter (Figure 2, D), air penetrates into the pore
system, at first into the lager pores. The capillary pressure “breaks through” locally. This
phenomenon has also been observed in inert materials. In soil mechanics, the pressure at the
start of air entry is referred to as air entry value [3].
The development of the capillary pressure and the air entry value are affected by several
influences like particle size distribution, evaporation rate and mobility of the particles.
Especially the amount of small sized particles like cement, fly ash or silica fume affects the
observed process because these particles form a narrow pore system which allows for smaller
menisci and higher pressure values.
Capillary pressure build-up and shrinkage can be numerically simulated on the particle
level [4]. This allows to study the effects of different influences on the observed phenomena.
In cement paste, mortar, concrete, and also in inert materials like soils, the capillary
pressure may be measured by using pressure sensors. Usually, a water filled tube connects the
pore system of the material to the actual sensor element. The use of such sensors for cement
based materials has been described by several authors, e.g. in [2] [5] [6].
The pressure “break-through” shown in Figure 2 is taking place when air reaches the
sensor tip. Because of material inhomogeneous, this is a local event. At other sensor positions,
the break-through might occur earlier or later. Therefore, the local break-through can not be
used as an indicator of reaching the air entry value. Previous experimental investigations have
shown that the air entry value may be identified by deformation and electric conductivity
measurements [7].
The aired pores are weak points in the system and the origin of strain localisation and
cracking. If the described air entry takes place at a time when the material has not reached a
significant strength yet, the cracking risk significantly increases.

3. CAPILLARY PRESSURE MEASSUREMENT UNDER SITE CONDITIONS

For measuring the capillary pressure under site conditions, optimized light-weight capillary
pressure sensors were built, see Figure 3 (left). They have a conic water filled sensor tip and
can be applied to the concrete surface after casting and compacting. The conic tip carries the
sensor’s weight and provides the hydraulic connection between the pore water in the material
and the sensor element. A cable connects the sensor with a digital recording device and
supplies it with power.
The cable connection, unfortunately, limits the usability of the capillary pressure sensors
under site conditions because the cables hinder surface finishing and, if they are moved, the
hydraulic connection between pore water and sensor may be interrupted. Therefore, capillary
pressure sensors with an integrated radio module were developed and prototypes have been
tested successfully. These wireless sensors connect automatically to a base station within a
radius of about 50 m and allow the monitoring of comparably large planar concrete structures
like roads or bridge decks.

Page 3
 Figure 3: Capillary pressure sensor (left), on-site capillary pressure measurement (right)

The cable based capillary pressure sensors have been used for several measurements under
site conditions. In Figure 3 the sensors may be seen applied to a real structure. The graph in
Figure 4 shows the measured development of the capillary pressure. The concrete had a
cement content of 335 kg/m³ and a water-binder ratio of 0.45. It contained air-entraining
admixtures. The capillary pressure sensors were applied shortly after surface finishing (60
minutes after casting). After about 100 minutes, the bleeding water on the surface was
evaporated and the absolute capillary pressure value started to rise. The average evaporation
rate amounted to about 0.5 kg/(m² h). It was measured by using a curing meter [8].

-80 0,0

Start of curing
Capillary pressure [kPa]

-60 (plasic foil)

-40 Sensor 1

-20 Surface
finishing Sensor 2
0

20 160,0

0 1 2 3 4 5 6 7 8
Time after casting [hours]

Figure 4: Capillary pressure development in concrete (left), small cracks on the surface (right)

Four hours after casting, the capillary pressure reached a value of about -45 kPa. At this
time the surface was covered with a plastic foil. As a result, the absolute capillary pressure
value temporarily decreased because the particles and the menisci on the surface were re-
arranged due to the contact with the foil. It can be seen that covering the concrete surface does
not prevent the continuing capillary pressure build-up which at this age is not only caused by
evaporation, but also by the beginning cement hydration.
In experiments performed under laboratory conditions, air entry values between -10 kPa
and -48 kPa were measured for cementitious materials [7]. It seems that for the structure

Page 4
shown in Figure 3 the curing was started too late since the capillary pressure amounted
already to -45 kPa. This would also explain why small cracks were found on the surface
before it was covered with the plastic foil, see Figure 4. Unfortunately, access to the structure
was limited in this case and no further investigations could be undertaken.
After removing the sensors, conic holes with diameters of about 1 cm at the surface and
depths of about 5 cm remain. They have to be closed by appropriate materials like cement
mortar.

4. AUTOMATED CONCRETE CURING

Experiments reported by Wittmann [2] [10] and Radocea [5] have shown that the capillary
pressure build-up can be reversed by adding water to the surface. After the added water is
evaporated, however, the absolute pressure value starts to rise again, but can be reduced by
repeatedly rewetting the surface. Figure 5 shows the capillary pressure development in a
rewetted concrete sample. It can be seen that the absolute pressure value drops down to zero
after rewetting the surface.

-80
Air entry at the sensor tip
Capillary pressure [kPa]

-60 Rewetting of
the surface
-40

-20

20
0 1 2 3 4 5 6 7 8
Time after casting [hours]

Figure 5: Capillary pressure development in a rewetted concrete specimen

On the basis of surface rewetting, an automated closed-loop controlled concrete curing

system was proposed [9]. An uncontrolled flooding of fresh concrete surfaces with water is
likely to cause damage to the surface layer and is not considered to be an applicable method.
Therefore, experiments with commercially available fogging equipment were undertaken, see
Figure 6 (left). Above the concrete surface, water is dispersed into very fine drops which tend
to sink down to the surface. The fog also increases the local air humidity. A very small
amount of water is sufficient to enlarge the radii of the menisci and to reduce the absolute
capillary pressure value.
The fogging system is connected to the capillary pressure sensors. Fogging starts
automatically when a defined capillary pressure is reached. The absolute value of this upper
threshold should be lower than the one of the material’s air entry value in order to prevent
plastic shrinkage cracking. For avoiding a closed water film on the surface of the concrete,
fogging is stopped when the capillary pressure reaches a lower threshold and resumed not
before the upper threshold is reached again. In this way, the capillary pressure is kept
automatically between two previously defined limits which prevent both capillary shrinkage
cracking and deterioration of the concrete surface.

Page 5
 Figure 6: Fogging device (left), concrete slab instrumented for controlled rewetting (right)

The closed-loop controlled curing system was tested under laboratory as well as under site
conditions. In a field experiment, two quadratic slabs with edge lengths of 2.70 m were
produced, see Figure 6 (right). They had a thickness of about 12 cm and were cast on a rough
concrete subgrade in order to restrain early age shrinkage. The concrete had an equivalent
water-binder-ratio of 0.47 and contained 290 kg/m³ of slag cement, 60 kg/m³ of fly ash, as
well as plasticizer.
The two slabs were cast simultaneously and instrumented with the capillary pressure
sensors shown in Figure 3. One of the slabs was cured by applying fog to its surface when the
capillary pressure had reached a threshold of -10 kPa. The second one was left uncured as a
reference sample. The average evaporation rate measured by a curing meter amounted to
approximately 0.34 kg/(m² h).

-60
Capillary pressure [kPa]

-50 Slab without curing

-40

-30

-20

-10

0
Cured slab
10
0 1 2 3 4 5 6 7 8
Time after casting [hours]

Figure 7: Capillary pressure versus time in a cured and in an uncured concrete slab

Figure 7 shows the capillary pressure versus time for the two concrete slabs. About 120
minutes after casting, the absolute capillary pressure value started to rise. After about 220
minutes, the pressure in the slab prepared for curing reached the predefined threshold of
-10 kPa and the rewetting with the fogging system started. It can be seen that the capillary

Page 6
pressure is not completely relieved by the rewetting. In the uncured slab, the capillary
pressure reached a value of about -53 kPa. After about 400 minutes, the curing system was
deactivated and the absolute capillary pressure value in the previously cured slab started to
rise rapidly. The rapid increase may be attributed to the narrower pore system in this age. It
was assumed that by this time the concrete had developed sufficient resistance to withstand
contracting forces caused by capillary pressure.

Figure 8: Crack pattern after 24 hours in a cured (left) and an uncured concrete slab (right)

After 24 hours, all cracks with widths larger than 0.1 mm were documented. Fig. 8 shows
the lengths and widths of the observed cracks. Unfortunately, the early age cracking risk for
this material was underestimated. Already before the curing was started, two cracks were
found in the surface of the slab which had been prepared for curing (at a capillary pressure of
about -9 kPa). Despite these unexpected cracks, it could be clearly demonstrated that the
capillary pressure controlled rewetting significantly reduces the early age cracking risk.

5. CONCLUSIONS
− Plastic shrinkage of concrete results from the build-up of a capillary pressure in the pore
system of the material and occurs mainly under high evaporation rates.
− The capillary pressure can be easily measured under site conditions with special
pressure transducers.
− On the basis of the measured capillary pressure, it is possible to make decisions
concerning the timing of curing measures and to evaluate the effect of such measures.
This allows to reduce the early age cracking risk.
− It appears to be possible to use the measured capillary pressure as feedback value for a
closed-loop controlled concrete curing based on surface rewetting. The capillary
pressure is automatically kept within certain limits in order to prevent early age damage.

Page 7
ACKNOWLEDGEMENTS
The financial support of the research by the German Federal Ministry of Education and
Research is gratefully acknowledged.

REFERENCES
[1] Slowik V., Schmidt M., Neumann A. and Dorow J., ‘Early age cracking and its influence on the
durability of concrete structures’, Proceedings of the 8th International Conference on Creep,
Shrinkage and Durability of Concrete and Concrete Structures - CONCREEP 8, 30 September -
October 2, 2008, Ise-Shima, Japan (Taylor & Francis Group London, 2008) Vol. 1, 471-477.
[2] Wittmann, F.H., ‘On the action of capillary pressure in fresh concrete’, Cement and Concrete
Research 6 (1976) 49-56.
[3] Peron, H., Laloui, L., Hueckel, T. and Hu, L., ‘Experimental study of desiccation of soil’. ASCE,
Geotechnical Special Pub 147 (2006) 1073–1084.
[4] Slowik, V., Hübner, T., Villmann, B. and Schmidt, M., ‘Capillary pressure and cracking in early
age concrete’, Proceedings of the International RILEM Symposium on Concrete Modelling -
CONMOD’08, May 26-28, 2008, Delft, The Netherlands (RILEM Publications SARL, 2008)
673-680.
[5] Radocea, A., ‘A study on the mechanism of plastic shrinkage of cement-based materials’,
PhD thesis, Chalmers University of Technology, Göteborg (1992).
[6] Hammer, T.A., 'The use of pore water pressure to follow the evolution from fresh to hardened
concrete', Proceedings of the 2nd International Symposium on Advances in Concrete through
Science and Engineering, 11-13 September, 2006, Quebec City, Canada (RILEM Publications
SARL, 2006).
[7] Slowik, V., Schmidt, M. and Fritzsch, R., ‘Capillary pressure in fresh cement-based materials and
identification of the air entry value’, Cement and Concrete Composites, 30 (7) (2008) 557-565.
[8] Jensen, O.M., ‘Monitoring water loss from fresh concrete’, Proceedings of the International
RILEM-JCI Seminar Concrete Durability and Service Life Planning -ConcreteLife’06;, Ein-
Bokek, Israel, 14-16 March, 2006 (RILEM Publications SARL, 2006) 197-202.
[9] Schmidt, D., Slowik, V., Schmidt, M. and Fritzsch, R., ‘Auf Kapillardruckmessung basierende
Nachbehandlung von Betonflächen im plastischen Materialzustand (Early age concrete curing
based on capillary pressure measurement)’, Beton- und Stahlbetonbau 102 (11) (2007) 789-796.
[10] Wittmann, F.H., ‘Heresies on shrinkage and creep mechanisms’, Proceedings of the
8th International Conference on Creep, Shrinkage and Durability of Concrete and Concrete
Structures - CONCREEP 8, 30 September - October 2, 2008, Ise-Shima, Japan (Taylor & Francis
Group London, 2008) Vol. 1, 3-9.

Page 8